Ion channels are transmembrane proteins that regulate entry of various ions into cells from the extracellular matrix. Ion channels are physiologically important, playing essential roles in regulating intracellular levels of various ions and in generating action potentials in nerve and muscle cells. Hille, B., Ionic Channels of Excitable Membranes (Sinauer, Sunderland, Mass., 1992). Passage of ions through ion channels is characterized by selective filtering and by a gating-type mechanism which produces a rapid increase in permeability. Angelides, K. J. and T. J. Nuttov, J. Biol. Chem. 258:11858–11867 (1981). Ion channels may be either voltage-gated, implying that current is gated (or regulated) by membrane potential (voltage), or chemically-gated (e.g., acetylcholine receptors and γ-aminobutyric acid receptors), implying that current is gated primarily by binding of a chemical rather than by the membrane potential. Butterworth, J. F. and G. R. Strichartz, Anesthesiology 72:711–734 (1980). An important characteristic of certain voltage-gated channels is inactivation: soon after opening they close spontaneously, forming an inactive channel that will not reopen until the membrane is repolarized. Miller, C., Science 252:1092–1096 (1991). Rapidly inactivating (“A-type”) voltage-gated ion channels control the rate at which excitable cells reach the threshold for firing action potentials and thus are key regulators of neuronal excitability. B. Hille, supra.
Many voltage-gated ion channels that generate action potentials have been cloned and sequenced, and all have a remarkably similar structure. A typical potassium channel contains four copies of an approximately 600-amino-acid polypeptide, each of which has six membrane-spanning α-helices. Heginbotham, L., et al., Science 258:1152 (1992). Sodium and calcium channels are single polypeptides of about 2000 amino acids that contain four homologous domains, each comprised of six transmembrane domains which are similar in sequence and structure to a potassium channel protein. These domains are connected and flanked by shorter stretches of nonhomologous residues. Jessell, T. M. and E. R. Kandel, Neuron 10(Supp):1–3 (1993). It is believed that the α-helical structures provide conformational flexibility for the ion channel which is functionally responsible for the channel's gating mechanism. See Heinemann, S., et at., J. Physiol. 88:173–180 (1994).
In addition to affecting action potentials, ion channels facilitate other important physiological functions such as cardiac pacemaking, neuron bursting, and possibly learning and memory. Crow, T., Trends Neurosci. 11:136–142 (1988); Hodgkin, A. L. and Huxley, A. F., J. Physiol. 117:500–544 (1952). In addition to their involvement in normal cellular homeostasis, ion channels are associated with a variety of disease states and immune responses. Diseases believed to be associated with dysfunction of ion channels include neurological disorders, metabolic diseases, cardiac diseases, tumor-driven diseases, and autoimmune diseases.
Due to the importance of ion channels in both normal cellular homeostasis and disease, considerable research effort has focused on ion channels, and particularly on identifying compounds which affect their function. Thus, several techniques have been developed to evaluate the gating mechanism of ion channels and the mode of action of channel-drug interaction. Electrophysiological recording has been used to define the roles of ion currents, and especially potassium and sodium currents, in generating action potentials in excised nerves. Hodgkin, A. L. and A. F. Huxley, supra. This technique, however, is not suitable for mass screening of compounds due to its technical complexity and the requirement of a high degree of sophistication to generate reproducible results. Radioligand binding assays have been used to characterize the site of action of various classes of ion channel blockers. However, the availability of radiolabelled ligands, the level of nonspecific binding, and the physico-chemical property of the molecules have limited the application of this technique. Strichartz, et al., Ann. Rev. Neurosci. 10:239–67 (1987). Fluorescent-labelled neurotoxin probes have also been used to map the molecular structure of the functional site of the channel, but have not gained general popularity for broader use. Angelides, K. A. and T. J. Nuttov, J. Biol. Chem. 256:11958–11967 (1983).
Recently, a modified yeast “two-hybrid” system has been developed to identify compounds that bind to either the NH2-terminal multimerization domain (commonly referred to as the “NAB” or “T1” domain) on the α-subunit of a Shaker-like potassium channel or to the “core” domain of the β-subunit of the potassium channel, thereby preventing the α/β subunit interaction. See U.S. Pat. No. 5,856,155 (M. Li), issued Jan. 5, 1999; and PCT App. No. PCT/US97/02292, published Aug. 28, 1997 (WO 97/31112). Unfortunately, significant inherent limitations in this system may prevent or limit its practical application. Such limitations include, for example, the extraordinarily tight nature of the α-NAB/β-core interaction (which survives such harsh treatments as detergent extraction and affinity chromatography), the limited applicability to potassium channels whose activity requires interaction between the NAB domain of the α-subunit and the core domain of the β-subunit, and, most importantly, the potentially significant inhibitory effect such compounds would have on potassium channel surface expression. [Regarding the tight association of α-and β-subunits, see Parcej, D. N., and J. O. Dolly, Biochem. J. 257:899–903 (1989) and Muniz, Z. M., et al., Biochemistry 31:12297–12303 (1992).] With respect to the latter limitation, β-subunits have been shown to promote N-linked glycosylation and surface expression of α-subunits. Shi, G., et al. Neuron 16:843–852 (1996). Thus, one would expect compounds that bind to the core domain of the β-subunit to block these chaperone-like effects, thereby reducing, if not preventing, the biosynthesis of functional potassium channels. By affecting the abundance or distribution of potassium channels in excitable membranes, such compounds would essentially act as ion channel blockers, and thus would likely have adverse neurophysiological effects. Finally, any compound that can effectively block the strong α-NAB/β-core binding interaction (i.e., compounds identified using this modified yeast two-hybrid system) must themselves have extremely high binding affinity for potassium channel subunits, and thus would likely be toxic to a mammalian host.
In view of the complexity of ion channel pharmacology and its attractiveness as a target site for the discovery of novel therapeutic compounds, there exists a need for an alternative technique which will enable the large-scale screening of compounds for ion channel modulatory activity in a simple and reliable manner. The present invention fulfills these and other needs.